Representativeness and Humanitarian Intervention

James Pattison

I. Introduction Humanitarian intervention involves military action by a state, coalition of states, or multinational organization with the primary purpose of preventing, reducing, or halting a mass violation of basic human rights. It is widely held that humanitarian intervention needs to meet certain criteria in order to be morally justiable. Most recent discussions of humanitarian intervention focus on the same few criteria, criteria derived from traditional Just War Theory (primarily the rules of jus ad bellum). These include the following. (1) The number of violations of basic human rights is large enough to justify humanitarian intervention. (2) There is a reasonable prospect of successfully tackling the humanitarian crisis. (3) The use of force is the last resort. (4) The intervener is a legitimate authority or has been authorized by a legitimate authority (which is typically taken to mean the United Nations Security Council). (5) The intervener has the right intent. (6) The intervener uses means to conduct the war that are consistent with its humanitarian aim. Much time and effort has been spent debating which of these criteria are morally important, why they are important, and what exactly they require.1 The preoccupation with Just War criteria has meant that two other factors in the legitimacy of humanitarian intervention have been, to some extent, overlooked. These two factors are more concerned with the views of those affected by intervention than their Just War counterparts, yet both signicantly affect the legitimacy of an intervener. The rst I shall describe as an interveners internal representativeness. This depends on whether an interveners decision making on the proposed intervention reects the opinions of its citizens. For instance, the internal representativeness of the 1992 American intervention in Somalia turned on whether America represented the opinions of Americans. The second is what I shall describe as an interveners external representativeness. This depends on whether an interveners decision making on the proposed intervention reects the opinions of those individuals in the political community that is subject to the intervention. To use the same example, the external representativeness of the 1992 American intervention in Somalia turned on whether America represented the opinions of Somalis. In this article, I make the case for the moral importance of these two factors that have been neglected in the literature to a certain extent.2 That is, I argue that an interveners legitimacy depends on whether it is representative of the opinions

JOURNAL of SOCIAL PHILOSOPHY, Vol. 38 No. 4, Winter 2007, 569587. 2007 Blackwell Publishing, Inc.

570

James Pattison

on intervention, rst, of its domestic population and, second, of those subject to its intervention. I begin by presenting three (largely complementary) arguments for the importance of an interveners internal representativeness. The rst is consequentialist: an intervener that has public support is more likely to be effective in tackling a humanitarian crisis. The second is the Resources Argument. This asserts that an intervener should be representative of its citizens opinions because these citizens provide the resources for humanitarian intervention. The third argument emphasizes the value of individual self-government. I then present three arguments for the importance of an interveners external representativeness. In some measure, these mirror the arguments for internal representativeness. The rst argument is consequentialist: an externally representative intervener is more likely to be effective. The second is the Burdens Argument, which holds that an intervener should represent the opinions of those subject to its intervention because those individuals are likely to be burdened by its intervention. The third argument again asserts the value of individual self-government. Overall, then, I present six arguments for the signicance of internal and external representativeness for the legitimacy of humanitarian intervention. Before we proceed, however, I need to clarify what I mean by representativeness. In her seminal work on the concept, Hanna Pitkin distinguishes between a number of meanings of representation, all based around the notion of re-presentation, a making present again.3 For example, formalistic views of representation include the authorization view, where a representative is someone who has been authorized to act, and the accountability view, where a representative is someone who is to be held to account.4 The problem with these views, Pitkin notes, is that they cannot tell us anything about what goes on during representation, how a representative ought to act, and whether he has represented well or badly.5 Alternatively, the descriptive view of representation takes representation to be standing for by virtue of a correspondence or connection between the representative and the represented. The focus is on the representatives characteristics, such as her class, ethnicity, and religion.6 But the most useful meaning of representation, at least for our purposes, is acting for. This view is concerned with the activity of representing, what goes on during representing, and the substance or content of acting for others.7 Accordingly, a representative institution will act for its citizens, by delegation or trusteeship. It is here that we nd the mandate-independence controversy. Should a representative represent his citizens opinions, since he is bound by mandate to do what they want, or should he have the independence to be able to promote his citizens interests as he sees them and as best he can? As will become apparent, in relation to humanitarian intervention I take the mandate side of this controversy. That is to say, a representative should represent his citizens opinions, a representative institution is one that reects its subjects opinions in its decision making, and representativeness is the measure of the extent to which an institution does so. It is also important to dene what I mean by an individuals opinions on the intervention. The most morally relevant opinion is an individuals view

Representativeness and Humanitarian Intervention

571

on whether humanitarian intervention should be undertaken. Other relevant opinionsbut largely secondary in importanceare an individuals views on the specic form of intervention (e.g., regime change or traditional peacekeeping), on who should intervene, and on how long the intervention should last. Those subject to a humanitarian crisis might want intervention, but not want it to be carried out by a particular intervener (such as the United States), or they might want regime change, but not long-term occupation. Furthermore, for reasons of practical simplicity (and perhaps of anti-paternalism), I am concerned with an individuals actual opinions rather than what his opinion would be if he had more information or if his opinion were more freely formed. Although individuals opinions may be inuenced in undesirable ways and contain misperceptions, I argue for their moral signicance when they relate to humanitarian intervention. II. Internal Representativeness Let us begin with the case for internal representativeness. To be internally representative, an intervener needs to reect, in its decision making, its citizens opinions on the proposed intervention. If the majority of its citizens do not want intervention, an internally representative government would not intervene. If its citizens want intervention to be undertaken in a particular way (such as regime change), then the decision making of the internally representative government would reect this.8 A would-be intervener can establish the opinions of its citizensand therefore be internally representativein a number of ways. For example, it could conduct opinion polls on a sample of the population, hold referenda on humanitarian intervention, and, less scientically, consider other indicators of the public mood, such as the media, its interactions with the public, and public campaigns. The latter sort of measures are, of course, not completely accurate, given media inuence, and, more generally, it can be tricky (but not impossible) to access reliable or genuine domestic public opinion. But an intervener should nevertheless attempt to garner such information, given the arguments that follow for the importance of internal representativeness. Note here that it is possible for nondemocratic states to be internally representative if they accurately reect their constituents opinions. That said, democratic states are perhaps most likely although far from certainto reect public opinion on intervention, given the democratic politicians desire to be elected, her sense of duty to reect her constituents opinions (and often public opinion more generally), and the likelihood of a concurrence between public opinion and the governments judgment. An immediate challenge might be this: Why does the question of internal representativeness for humanitarian intervention arise? On many issues (such as health, education, and scal policy), it seems right that elected politicians should have some independence to use their judgment. They should primarily act in accordance with what they deem to be in the national (or their constituents) interest, without always having to reect public opinion. In other words, the

572

James Pattison

trusteeship conception of representation according to which a representative can go against constituents declared opinions and use their own judgment seems appropriate in many contexts. Why should we prefer a delegate conception of representation according to which a representative must reect the opinions of their constituents in the context of humanitarian intervention? What distinguishes humanitarian intervention from other governmental acts such that it requires politicians to reect their citizens opinions? There are two distinctive features. The rst is that humanitarian intervention is a different sort of governmental action because it is not (usually) in the interests of the citizens of the intervening state. The trusteeship model of representation holds that representatives should have the freedom to promote the interests of their citizens (or constituents). However, since humanitarian intervention is not (usually) in the interests of the interveners citizens (or constituents), it transcends the remit of representatives on this model. To put this in another way, if we view the primary role of government as the promotion of its citizens interests, it follows that government contravenes its duciary obligation to its citizens by undertaking humanitarian intervention. We need to tread carefully here, however. In particular, we need to avoid endorsing a similar, but stronger, viewwhat Allen Buchanan, in his discussion of the internal legitimacy of humanitarian intervention, terms the discretionary association view of the state.9 This view understands the state as:
the creation of a hypothetical contract among those who are to be its citizens, and the terms of the contract they agree on are justied by showing how observance of these terms serves their interests. No one elses interests are represented, so legitimate political authority is naturally dened as authority exercised for the good of the parties to the contract, the citizens of this state.10

Accordingly, government is taken to be solely the agent of the associated individuals and its role as the furthering of these individuals interests. Indeed, on this position government acts legitimately only when it occupies itself exclusively with the interests of the citizens of the state of which it is the government.11 The problem with this view, as Buchanan points out, is that it is too strong.12 It denies that government possesses any obligations to those beyond the borders of the state. It follows that, on the one hand, almost any action (e.g., imperialism, colonization, and exploitation) could be justied on this view if it would advance the interests of those within the state, regardless of the harm caused to those beyond its borders. On the other hand, it also follows that any governmental action that is not in its citizens interests, such as the removal of unfair trade barriers, is unjustiable.13 Nevertheless, the notion that the specialness of humanitarian intervention arises from the breaking of the intervening governments duciary obligation to its citizens seems plausible. But we need to be clear about the strength of this obligation. It is not absolute: this is demonstrated by the inadequacies of the

Representativeness and Humanitarian Intervention

573

discretionary association view. Rather than holding that government acts legitimately only when it occupies itself exclusively with the interests of its citizens, we can say that the primary role of government is to promote its citizens interests. By viewing this duciary obligation as primary, this more moderate approach allows room for government to possess certain obligations to those beyond its borders, for instance, to avoid causing large-scale environmental pollution in a neighboring state. Yet, on this approach, these obligations are limited, given the primary role of government. And, as a substantial undertaking, humanitarian intervention seems to go beyond the scope of governments limited obligations to those beyond its borders and is incongruous with governments duciary obligation to its citizens.14 So, unlike the discretionary association view, this more moderate approach can admit that government possesses some limited obligations to those beyond its borders, but like the discretionary association view, it holds that, by undertaking humanitarian intervention, government contravenes its duciary obligation. On its own, however, the suggestion that humanitarian intervention is a special case because of its generally altruistic character is incomplete. If we limit the specialness of humanitarian intervention to only its (apparent) altruism, any humanitarian intervention which is in the interests of the citizens of the intervening state can be left to representatives to decide independently. We therefore need to identify a second feature that distinguishes humanitarian intervention from other governmental actions and means that we should reject a trusteeship conception of representation in this context. My suggestion is that what differentiates humanitarian intervention, in addition to its (apparently) altruistic character, is that it involves the use of military force and, more generally, extremely high moral stakes. Humanitarian intervention (like any use of military force) has signicant potential to cause high levels of suffering and devastation to those in the target state, for instance, by killing innocent civilians, destroying vital infrastructure, and creating a power vacuum. Yet a governments decision to undertake humanitarian intervention can also have considerable positive benets, such as protecting populations from genocide and ethnic cleansing. It follows that the consequences, either good or bad, of an agents decision if, when, and how to undertake humanitarian intervention will be considerable for those suffering the humanitarian crisis. Furthermore, the effects of an agents decision to intervene reverberate around the international system, not only by affecting international norms (both legal and normative) by, for instance, the setting of precedents, but also more materially by, for instance, creating refugee ows and destabilizing surrounding regions. The interveners citizens are also affected by the decision, for (as discussed below) they provide the nancial and human resources (which can be signicant). As such, the consequences, either good or bad, of an agents decision to intervene will also be highly signicant for the international system and the interveners citizens. So when making a decision that involves the use of military force and, more generally, has such high moral stakes, it seems right that a government should reect its citizens opinions in its decision making. Unlike for other, less

574

James Pattison

important, decisions, which we can leave politicians to decide for themselves, trusteeship is not appropriate when the moral stakes are so high. Hence, there are two elements to the specialness of humanitarian intervention: First, humanitarian intervention tends to go against the intervening states primary (but not absolute) obligation to its citizens; second, it involves the use of military force and, more generally, high moral stakes. Although we have two reasons why trusteeship is unpersuasive in the context of humanitarian intervention, we do not yet have justication for why the delegate conception of representation should be preferred. In other words, we now need to see why an interveners decision making on the proposed intervention should reect the opinions of its citizens. In what follows, I present three arguments for the importance of internal representativeness for the legitimacy of humanitarian intervention. Note here that for the rest of the article I will use representativeness in the sense of representation as delegation. Also note that, although these three arguments could be applied to make the case for a delegate conception of representativeness for other governmental decisions, they are particularly pertinent for humanitarian intervention, given the two distinctive features outlined. For instance, one option would be to apply these three arguments to make the case for the representativeness of decisions that have lower moral stakes. However, the fact that these other decisions have lower moral stakes means that these arguments would not be as persuasive as they are for humanitarian intervention. It is less important, for instance, that there is individual self-government on the issue of public transport than on the issue of humanitarian intervention. (i) Increased Effectiveness Let us begin the case for the importance of internal representativeness with a consequentialist argument. One of the largest problems faced by humanitarian intervention is insufcient commitment. This has led to critically under-resourced, and ultimately unsuccessful, interventions. The failure of UN member states to provide UNAMIR, the UN force led by Romo Dallaire, with the necessary resources to stop the genocide in Rwanda is the most conspicuous example. Many of these problems arise because interveners are unwilling to commit the necessary nancial, military, and diplomatic resources to potentially unpopular and controversial interventions. By contrast, an internally representative intervener which knows that it has public support is more likely to be willing to commit the resources required to be successful. It may be more willing, for instance, to risk casualties and so be able to undertake ambitious military maneuvers, which are necessary for intervention to be successful. Consider, in this context, Australias 1999 action in East Timor. Since it knew it had the support of the Australian public, the Australian government was prepared to accept some casualties and, as a result, intervened with the level of military force necessary for successful humanitarian intervention.15

Representativeness and Humanitarian Intervention

575

This consequentialist argument for internal representativeness is, however, contingent on there being a correlation between internal representativeness and effectiveness. On occasion, being internally representative may not ensure that the intervention is successful. The time it takes to establish whether there is public support for intervention may mean that deployment is slowed, which in turn undermines the effectiveness of the operation. Alternatively, public opinion may change during the intervention, but if the intervener were to respond to this change (perhaps by altering its mission objectives), it would be less effective. Likewise, humanitarian intervention can be successful without being internally representative. This raises an important question for both internal and external representativeness: Would an intervener be legitimate if it lacked internal or external representativeness (or both), yet was likely to be effective at preventing, reducing, or halting the mass violation of basic human rights? The answer depends on the circumstances, and in particular, on how effective the intervener would be at tackling the mass violation of basic human rights. If the benecial consequences were enormous (such as the prevention of genocide), an intervener that lacked internal or external representativeness (or both) would lose some legitimacy, but it might still be legitimate overall. Suppose, for example, if in the beginning of the mass slaughter in Rwanda in early 1994, Uganda had been willing to intervene and expected to do so effectively. Given that this could have saved hundreds of thousands of lives, the fact that Uganda was undemocratic at the time, and might not have consulted with the Rwandans, may not have undermined the overall legitimacy of Ugandas intervention. Conversely, if the likely benecial consequences of the interveners action were not as substantial (although still serious enough for intervention to be warranted), then the intervener would need to be internally and externally representative in order to be legitimate. As such, the likely prevention of a smallerbut still signicantnumber of violations of basic human rights (such as the removal of an oppressive regime) cannot trump the importance of the intervener representing both the opinions of its citizens and those subject to its intervention. That is not to deny that internal and external representativeness are important considerations for the legitimacy of humanitarian intervention. On the contrary, the six arguments set out in this article establish that these are clearly signicant considerations. The aim here is to make clear the strength of the arguments that follow and, in particular, to avoid overstating the case. To clarify, then, the exact signicance of these considerations: internal and external representativeness carry considerable weight in the assessment of an interveners legitimacy. They are not, however, sufcient conditions for legitimacy: other factors, such as the effectiveness of the intervention in tackling the humanitarian crisis and the means used, are also important. Nor are they necessary criteria for legitimacy: on occasion, their importance can be outweighed by highly benecial consequences. That said, in most cases of humanitarian intervention, where extremely benecial consequences are not on the cards, an intervener would be illegitimate if it lacks either internal or external representativeness. As such, in most cases, internal and

576

James Pattison

external representativeness should be regarded as necessary conditions for legitimate humanitarian intervention. (ii) The Resources Argument Having determined how internal and external representativeness relate to consequentialist thinking on humanitarian intervention, I will now consider the second reason why an interveners internal representativeness matters. This is what I call the Resources Argument. The central contention is this: since the interveners citizens provide the resources for humanitarian intervention, their opinions should be reected in the decision on intervention. The underlying argument at work here is Lockean: an individual should have some freedom to determine how his own resources (property) are used. Given that humanitarian intervention requires a substantial amount of resources, the intervener should reect the opinions of those providing the resources for humanitarian interventionits citizens. Doing so means that these individuals retain some control over their resources. This Lockean argument is not absolute. There are moral constraints on how an individual should use his resources (such as not causing excessive harm to others) and the importance of individual choice here might not be as signicant as other moral considerations (such as highly benecial consequences). Nevertheless, some degree of control over ones own resources is intuitively attractive. In theory, we could make this argument about any governmental action that uses its citizens resources. However, it is more convincing for humanitarian intervention because of the level of resources involved. Alex de Waal estimates that the Economic Community of West African States (ECOWAS) interventions in Liberia and Sierra Leone cost $4 billion.16 In The Responsibility to Protect, the International Commission on Intervention and State Sovereignty estimated that the cost of the Kosovo intervention (including post-intervention peacekeeping and reconstruction) was $48 billion.17 The interveners citizensin these cases, the citizens of ECOWAS and NATO respectivelyultimately have to foot the bill for humanitarian intervention, perhaps through signicantly increased taxation or greatly decreased public spending elsewhere. It is right, therefore, that these individuals should have some input into the decision making on humanitarian intervention. The Resources Argument gains further plausibility if, in addition to nancial resources, it includes human resources. The interveners citizens provide the personnel to undertake humanitarian intervention. Some of these individuals may be injured and killed in combat. There is further reason then for representing the opinions of these individuals.18 (iii) Individual Self-Government The Resources Argument is persuasive as far as it goes, but it does not go far enough. It does not quite capture the main reason why an interveners internal

Representativeness and Humanitarian Intervention

577

representativeness matters: individuals should have some control over their governing institution because it is their governing institution. More specically, the citizens of the intervener should have their opinions on the intervention represented because it is their intervener: it is their state or their multinational organization that is intervening. This sentiment was discernible in the early stages of the 2003 war in Iraq; many protesters in the United Kingdom claimed that the war was conducted not in our name.19 Their protest was not about the use of resources; it was against the fact that their government was undertaking an action which they opposed. Accordingly, I will now outline a third, more Rousseauian defense of the importance of an interveners internal representativeness. This third argument relies on the principle of individual self-government, which runs as follows: a governing institution should reect the wishes of its citizens such that it is as if those individuals were in authority themselves. Individual self-government here possesses signicant value. In Robert Dahls words: To govern oneself, to obey laws that one has chosen for oneself, to be self-determining, is a desirable end.20 To be sure, individual self-government is not always an overriding value; rather, more individual self-government is by and large desirable. Occasionally, other moral factors (such as highly benecial consequences) may trump the importance of individual self-government, but this is not to deny its value. Indeed, individual self-government seems to possess nonconsequentialist value. The fullment of an individuals wishes on how he wants to be governed is valuable regardless of whether these wishes, if they were realized, would contribute to his well-being. To see this, consider a (hypothetical) society whose government is hierarchical and unrepresentative. It never consults its citizens on how they wish to be governedit makes decisions by decreebut is competent at promoting the interest of its citizens. Although such a government would not be that morally objectionable because it would be promoting its citizens interests, something morally important is still missing. That missing element is the value of individuals having a signicant input in how they are governed and how their society is run. To be fully compatible with the principle of individual self-government, both the structure of government and every law it makes would need to match each individuals opinions on how they wish to be governed. Yet, in all but the smallest of societies, complete self-government is unachievable. This is what Thomas Christiano calls the incompatibility problem.21 Given the inevitable conict of opinions that arises in a society, the ability of a number of individuals to choose how they are governed will be frustrated. But this does not mean that the importance of an institution representing the opinions of its citizens cannot be justied by the principle of individual self-government. The crucial point is that we are not concerned solely with achieving full individual self-government within a society (which is a chimera), but with increasing the amount of individual selfgovernment. As such, we are concerned with the relative, rather than absolute, level of individual self-government. It follows that an intervener that represents at least the majority of its citizens opinions on the humanitarian intervention is

578

James Pattison

likely to have more individuals who are self-governing on this issue than an intervener that does not. For instance, requiring a super-majority (say of twothirds of the voting population) for intervention would risk giving those who oppose intervention a greater say than those that support itand therefore decrease the overall amount of individual self-government on the issue of humanitarian intervention.22 The value of individual self-government has a considerable impact on the argument for an interveners internal representativeness. An interveners internal representativeness is morally signicant because of the importance of individuals having a voice in the running of their political institutions. As a signicant undertaking by the state, it is important that humanitarian intervention be responsive to the concerns of individual self-government by being representative of its citizens opinions on intervention. An individuals freedom to choose whether there should be intervention, who should do it, how long it should last, and what form it should take, therefore matters. Indeed, given the non-instrumental value of individual self-government, there is reason for an intervener to be internally representative even if its population is mistaken on the issue of intervention. To be sure, this reason may not always be decisive. There may be more morally urgent concerns, such as the likely achievement of extremely benecial consequences. Nevertheless, individual self-government is a central reason why an interveners internal representativeness matters. An intervener should be internally representative and respond to its citizens opinions because those are the opinions of its citizens. Suppose, for example, that the Vietnamese government were considering intervening in Laos to tackle a (hypothetical) genocide. The Vietnamese people supported intervention, but only with UN Security Council authorization. The views of the Vietnamese people matter, on the logic of this self-government argument, neither because taking account of those views will best serve international law and order, nor because doing so is the best for the Vietnamese peoples enjoyment of basic human rights, but because it is their state, Vietnam, that is considering intervening. It may help to summarize briey the argument thus far. The rst reason for the signicance of an interveners internal representativeness is consequentialist: an internally representative intervener is more likely to be effective. The second is the Resources Argument, which asserts that the interveners citizens should have their opinions reected in its decision making since they provide the resources for humanitarian intervention. The last reason is the value of individual self-government on humanitarian intervention. Together, these three reasons demonstrate that an interveners internal representativeness is an important consideration for (and usually necessary condition of) the legitimacy of humanitarian intervention. III. External Representativeness To be externally representative, an intervener needs to represent the opinions of those in the political community that is potentially subject to its humanitarian

Representativeness and Humanitarian Intervention

579

intervention. For instance, an externally representative intervener would not undertake humanitarian intervention if those who would be subject to it do not want intervention. Similarly, if those individuals do not want a particular form of intervention, the decision making of the externally representative intervener would reect that. To establish the opinions of those subject to its humanitarian intervention, a would-be intervener should, rst, attempt to obtain direct access to these individuals. Sometimes there are obstacles to achieving this, but these are not always insurmountable. The International Committee of the Red Crosss (ICRC) People on War survey, for instance, comprised a series of comprehensive opinion polls and interviews on humanitarian intervention in a number of war-affected states. Among the ndings was that sixty-six percent of those surveyed wanted more intervention from the international community to deal with humanitarian crises, and only seventeen percent wanted less.23 In addition, they were able to distinguish between combatants and civilians, as well as to identify those who had suffered severe burdens caused by conict. Of course, such useful information will not always be accessible before the launch of a humanitarian intervention. Access may be denied to researchers and the situation may be too dangerous (the ICRC also faced these difculties with its research). Where direct consultation with those suffering the humanitarian crisis is impossible, an externally representative intervener will not simply presume these individuals opinions on the proposed intervention. Instead, it will use secondary sources or indicators of these citizens opinions, provided, for instance, by intermediaries. The challenge for the intervener, if it is to be externally representative, is to nd reliable agents that provide accurate information on the opinions of the victims and affected bystanders. One way that the intervener can determine whether an agent provides accurate information is by examining its ethos, track record, and agenda. Another way is to compare the agents account with that of the few citizens with whom direct access is possible (e.g., refugees). The agents that are perhaps most likely to be reliable are certain non-governmental organizations and what Mary Kaldor calls islands of civility (groups that have political support but are not involved in the violence).24 An intervener, therefore, can be externally representative in a number of ways. Although these are not always easily achieved, in what follows I argue that an intervener should make a concerted effort to be externally representative. A signicant part of its legitimacy depends on its doing so. This is the case even if Jacques deLisle is right in asserting that most victims will not oppose intervention.25 It is important to establish that this is true: that those subject to the humanitarian crisis clearly want intervention. Indeed, much of the opposition to humanitarian intervention revolves around the idea that it is paternalistic, forced upon people who do not want it.26 One logical corollary of this objection is that, if intervention is to be justiable, the interveners external representativeness is vital. As Fernando Tesn notes, leaders must make sure before intervening that they have the support of the very persons they want to assist.27

580

James Pattison

Yet, the question remains: why exactly is it that the intervener should establish and represent the opinions of those in the political community that is subject to its intervention? The three reasons for the importance of an interveners external representativeness mirror to a certain degree the three reasons presented for internal representativeness. The rst claims that an externally representative intervener is more likely to be effective. The second is the Burdens Argument, which asserts that those subject to the humanitarian intervention should have their opinions represented because intervention is likely to burden them. The third emphasizes the value of individual self-government. Together, these three arguments will show that external representativeness is a signicant factor in (and, apart from extreme cases, a necessary condition of) the legitimacy of an intervener. Indeed, external representativeness perhaps carries greater weight in the overall assessment of an interveners legitimacy than internal representativeness. This is because we can expect the three arguments presented to be even more signicant in this context: (1) An interveners external representativeness is likely to be of greater consequence for its effectiveness than its internal representativeness; (2) the burdens of intervention are likely to have a larger effect on individuals than the contribution of resources to undertake intervention; and (3) individual self-government seems to be even more important when you are subject to intervention than when undertaking it. (i) Increased Effectiveness I begin the defense of the moral importance of external representativeness with a plainly consequentialist argument: an intervener that represents the opinions of those subject to its humanitarian intervention is more likely to be effective at preventing, reducing, or halting the mass violation of basic human rights. This is because, rst, prior consultation with those who would be subject to intervention can indicate whether there is widespread support for intervention in the target state. This is a key factor determining whether intervention will succeed.28 The ofcial British peacekeeping manual thus states:
Without the broader co-operation and consent of the majority of the local population and the leadership of the principal ruling authorities, be they party to the dispute or government agencies, success is not a reasonable or realistic expectation. . . . Put simply, consent (in its broadest form) is necessary for any prospect of success.29

Without such prior consultation, the intervener might undertake action that is unpopular with the local population and, as a result, face high levels of resistance, making successful intervention difcult. In addition, an externally representative intervener is more likely to know whether a particular course of action or mission during the intervention will be successful. In this context, Jarat Chopra and Tanja Hohe assert that locals tend to have the best knowledge of the situation, including, we can surmise, the location of conict hotspots, the terrain and weather conditions, and the underlying political factors.30 By consulting with locals, therefore,

Representativeness and Humanitarian Intervention

581

an intervener will have a greater awareness of this situation and, consequently, will be better placed to undertake successful intervention. Moreover, since an externally representative intervener reects, in its decision making, the opinions of those subject to its intervention, it is more likely to make these individuals feel involved with the intervention. Theodora-Ismene Gizeles and Kristin Kosek argue that this feeling of involvement is necessary for effective intervention.31 Conversely, a population that is largely uninvolved in humanitarian intervention is less likely to cooperate with the intervening parties or expend efforts to make the intervention successful.32 (ii) The Burdens Argument The second argument for external representativeness is what I call the Burdens Argument. This asserts that an intervener should represent the opinions of those in the political community that is subject to its intervention because of the potential burdens imposed by humanitarian intervention. Those in this community might have to suffer civilian and military casualties, damage to vital infrastructure, increased levels of insecurity, and other costs associated with being in a war zone. Given that these individuals face these burdens, it is important that the intervener should reect their opinions on the intervention. This Burdens Argument is similar to the Resources Argument for internal representativeness in that it relies on the importance of individual choice. Whereas the underlying principle of the Resources Argument is that an individual should have some choice over how his resources are used, the underlying principle of the Burdens Argument is that an individual should have some choice over the burdens he faces. The reason why individual choice regarding burdens matters is that those suffering burdens are negatively affected. More precisely, a burden of humanitarian intervention can be dened as a negative impact on an individuals basic human interests caused by that intervention. Examples of burdens therefore include injury, disruption of food supplies, and damage to vital infrastructure (e.g., basic medical services and running water). Hence, the Burdens Argument holds that those subject to the humanitarian intervention should have their opinions represented because intervention may have a negative impact on their basic human interests.33 As it stands, this Burdens Argument is both too inclusive and too exclusive. It is too inclusive because it suggests that the intervener should reect the opinions of all those in the political community that is subject to its intervention. This includes the opinions of those carrying out the violations of basic human rights, which create the need for intervention. For example, on the logic of this argument, NATO should have represented the opinions of the leaders of the Bosnian Serbs before undertaking its air strikes in 1995, since they were essentially the targets and were burdened by this action. We therefore need to amend the Burdens Argument to take into account moral culpability. In this context, Tesn asserts that it is the victims of the oppression who must welcome intervention.34 More specically, he argues:

582

James Pattison

[I]n a tyrannical regime the population can be divided into the following groups: the victims; the accomplices and collaborators; and the bystanders. . . . Of these groups, only the rst, the victims, have (arguably) a right to refuse aid. The accomplices and bystanders who support the regime are excluded for obvious reasons. Their opposition to intervention does not count. And the bystanders who oppose the regime cannot validly refuse foreign aid on behalf of the victims.35

Although this typology is illuminating, it is too simplistic. I agree that the opinions of accomplices and collaborators should be given no weight. Any burdens of intervention they suffer are a consequence of their own morally reprehensible behavior. I also agree that we should assign greatest weight to the opinions of the victims. They are not usually morally culpable for the humanitarian crisis, yet often face some of the largest burdens of intervention, such as the bombing campaigns conducted in the regions in which they live. Moreover, if a potential intervener treats each individuals opinions equally, and if the majority of others (such as the bystanders) oppose humanitarian intervention, the victims would be left to suffer the humanitarian crisis. For this reason, we should give most weight to the opinions of the victims. Yet I disagree with Tesns rejection of the importance of the bystanders opinions. Although they are less important than the opinions of victims, some bystanders opinions should be represented as well. In particular, we should include the opinions of those bystanders who are likely to be burdened by the intervention precisely because they are burdened bystanders: they are not (directly) responsible for the humanitarian crisis but might suffer in its resolution. Hence, an externally representative intervener will, rst, give most weight to the opinions of the victims of the humanitarian crisis and, second, take into account the opinions of the bystanders likely to be burdened by the intervention. Of course, it is not always easy to distinguish between victims, bystanders, collaborators, and accomplices.36 But, although sometimes the line between the victims and the aggressors is blurred, on other occasions it is all too apparent who are the victims and who are the aggressors. As it stands, this Burdens Argument is also too exclusive. Some of the burdens of humanitarian intervention may fall on those outside the borders of the target state. An obvious example is the creation of a refugee ow that destabilizes a neighboring state. Therefore, we need to amend the Burdens Argument so that, when individuals in other political communities will be burdened by the interventionwhen they will also be burdened bystandersthe intervener gives some weight to their opinions too. That said, in most cases, the effects on those beyond the borders of the target state would not be signicant enough to warrant the consideration of these individuals opinions. (iii) Individual Self-Government Like the Resources Argument in relation to internal representativeness, the Burdens Argument does not provide a complete defense of the importance of an

Representativeness and Humanitarian Intervention

583

interveners external representativeness. That is, it does not encapsulate fully why an intervener should be externally representative. For this, we need to turn to the third argument for external representativeness, which invokes the value of individual self-government.37 Let us start with the instrumental argument for individual self-government in this context (and, by implication, for an interveners external representativeness). This instrumentalist justication relies on a form of what Albert Weale terms the non-paternalist principle.38 To be specic, individuals are the best judge of what enhances their well-being in most cases, although there are obvious exceptions.39 Individual self-government is valuable, therefore, because self-governing individuals are more likely to realize their well-being. It follows that an institution that is representative, in that it reects its citizens opinions in its decision making, is more likely to promote its citizens well-being.40 It also follows that an intervener that represents the opinions of those subject to its interventionand is therefore externally representativeis more likely to promote (or, at least, not harm) these citizens well-being. This is because the intervener, by reecting these individuals wishes, desires, and goals in its decision making, will help them to attain what they themselves identify as being required for their well-being. For instance, suppose an intervener responds to a humanitarian crisis in a society which has strong religious customs. These customs form part of what constitutes the good life for many individuals. By consultation, an externally representative intervener would learn that these religious customs and practices contribute to many individuals well-being in this society. It would therefore have a better understanding of what is necessary to promote these individuals well-being. It might follow, for example, that the intervener involves religious leaders in a transitional administration and avoids damaging religious buildings. I argued earlier that individual self-government matters in itself: it is important that an individual should be self-governing even if his opinions, if realized, would not obviously promote his well-being. This non-consequentialist value of individual self-government adds to the importance of external representativeness. A state, coalition of states, or multinational organization should not intervene to protect those who do not want their political community to be subject to humanitarian intervention. This is the case even if intervention would promote these individuals well-being in the short term, for instance, by protecting them from being the victims of oppression and from the violation of their basic human rights.41 Moreover, it is not only individuals opinions on whether there should be intervention that matter for the representativeness of an intervener. Although this tends to be the most prominent issue, it also matters that an intervener responds to other opinions of those subject to its intervention, including their opinions on who should intervene, on the form intervention should take, and on how long it should last. The opinions of those subject to the intervention on these issues also have value. For instance, those subject to a humanitarian crisis might desire intervention, but have grievances against the proposed intervener.42 Responding to such

584

James Pattison

grievances might not directly promote the well-being of those subject to the interventionan alternative intervener might not be any more effectivebut it is still important to be responsive to these opinions as a matter of individual selfgovernment and, ultimately, external representativeness.

IV. Conclusion The principal purpose of this article has been to highlight, and to make the case for, the moral signicance of two largely overlooked factors for the legitimacy of humanitarian intervention: whether the intervener is representative of the opinions, rst, of its citizens and, second, of those in the political community in which it intervenes. There are three, largely complementary, reasons why the rst factor, the interveners internal representativeness, is important. The rst is consequentialist: an internally representative intervener is more likely to be effective because it is more likely to commit the resources necessary for successful humanitarian intervention. The second is the Resources Argument: the intervener should take into account its citizens opinions on the intervention because its citizens provide the nancial and human resources for intervention. The third is the value of individual self-government on humanitarian intervention. Three parallel reasons explain the importance of the second factor, the interveners external representativeness. The rst is consequentialist: an externally representative intervener is more likely to be effective. The second is the Burdens Argument: the intervener should take into account the opinions of those in the political community in which it intervenesand, in particular, the opinions of the victims and burdened bystandersbecause humanitarian intervention may have a negative impact on these individuals basic human interests. The third is the value of individual self-government. Hence, internal and external representativeness play a signicant role in an interveners legitimacy. For that reason, we need to pay them greater attention and, ultimately, to improve the extent to which current interveners are internally and externally representative, in addition to the meeting criteria derived from traditional Just War Theory. And although it can be difcult for an intervener to obtain accurate information on the opinions of both its constituents and those suffering the humanitarian crisis, these difculties are not insurmountable. For the reasons given in this article, an intervener should make a concerted effort to obtain and take into account such information and consequently be both internally representative and externally representative. Earlier versions of this article were presented to the Newcastle Political Philosophy Group, the Pavia Conference in Political Philosophy, and the Annual Conference of the Global Studies Association in 2005. I would like to thank, in particular, Derek Bell, Thom Brooks, Simon Caney, Ian Carter, Peter Jones, Graham Long, Ian OFlynn, and two anonymous reviewers for their helpful comments.

Representativeness and Humanitarian Intervention

585

Notes
1

The most prominent example of this approach is the report of the International Commission on Intervention and State Sovereignty (ICISS), The Responsibility to Protect (Ottawa: International Development Research Centre, 2001). This argues that humanitarian intervention needs to have a just cause, meet four precautionary principles (right intention, last resort, proportional means, and reasonable prospects), and be authorized by the appropriate legal authority. Other notable examples include Tom Farer, Cosmopolitan Humanitarian Intervention: A Five-Part Test, International Relations 19, no. 2 (2005): 21120 and Nicholas Wheeler, Saving Strangers (Oxford: Oxford University Press, 2000). For a rejection of the moral importance of legal authority, see my Humanitarian Intervention and International Law: The Moral Signicance of an Interveners Legal Status, Critical Review of International, Social and Political Philosophy 10, no. 3 (2007): 30119. 2 That said, these issues (especially external representativeness) have received some treatment. The most constructive discussions are Allen Buchanan, The Internal Legitimacy of Humanitarian Intervention, The Journal of Political Philosophy 7, no. 1 (1999): 7187; Allen Buchanan, Justice, Legitimacy, and Self-Determination: Moral Foundations for International Law (Oxford: Oxford University Press, 2004), 98105; Allen Buchanan, Institutionalizing the Just War, Philosophy and Public Affairs 34, no. 1 (2006): 27; Jacques deLisle, Humanitarian Intervention: Legality, Morality, and the Good Samaritan, Orbis 45, no. 4 (2001): 552; ICISS, Responsibility to Protect, 36; David Luban, Intervention and Civilization: Some Unhappy Lessons of the Kosovo War, in Global Justice and Transnational Politics: Essays on the Moral and Political Challenges of Globalization, ed. Pablo De Greiff and Ciaran Cronin (London: MIT Press, 2002), 8486; Fernando Tesn, Humanitarian Intervention: An Inquiry into Law and Morality, 2nd ed. (New York: Transnational Publishers, 1997), 12629; and Fernando Tesn, The Liberal Case for Humanitarian Intervention, in Humanitarian Intervention: Ethical, Legal and Political Aspects, ed. J. L. Holzgrefe and Robert O. Keohane (Cambridge: Cambridge University Press, 2003), 1057. 3 Hanna Pitkin, The Concept of Representation (London: University of California Press, 1967). 4 Ibid., 3855. 5 Ibid., 58. 6 Ibid., 6191. 7 Ibid., 113. 8 One example of an internally representative humanitarian intervention was the 1999 UN-authorized, Australian-led, intervention in East Timor. In response to images of the atrocities on their television screens, the Australian public opinion was crucial to providing the impetus for the Howard governments intervention. See Nicholas Wheeler and Tim Dunne, East Timor and the New Humanitarian Interventionism, International Affairs 77, no. 4 (2001): 80527. 9 Buchanan, Internal Legitimacy, 7187. 10 Ibid., 7475. 11 Ibid., 75, emphasis added. 12 Ibid., 78. 13 Moreover, even if a humanitarian intervention receives a special democratic mandate, it would not be legitimate on the strict logic of the discretionary association view. The democratic mandate would have to be unanimous (which is improbable); otherwise, humanitarian intervention would violate the terms of the contract for those who oppose intervention. See ibid., 76. 14 This is not to deny that individuals within the state possess extensive obligations to those beyond their borders. Indeed, these obligations could plausibly include a duty to encourage and support humanitarian intervention. This is consistent with the notion that the primary role of government is to promote its citizens interests. 15 In fact, Australia successfully intervened without suffering any casualties, even though it had braced itself for them. See Wheeler and Dunne, East Timor, 80527.

586
16

James Pattison

Alex de Waal, Who Fights? Who Cares? War and Humanitarian Action in Africa (Eritrea: Africa World Press, 2000), 81. 17 ICISS, Responsibility to Protect, 71. 18 This concern about resources relates to the rst argument for increased effectiveness. If citizens agree to use their resources for humanitarian intervention, the intervener is more likely to be successful. By contrast, if they generally oppose the nancial and human costs of humanitarian intervention, the intervention is less likely to be successful. 19 In using this example, I am not claiming that the war on Iraq was a humanitarian intervention. 20 Robert Dahl, Democracy and Its Critics (London: Yale University Press, 1989), 89. 21 Thomas Christiano, The Rule of the Many: Fundamental Issues in Democratic Theory (Oxford: Westview Press, 1996), 25. 22 Suppose a minority of two-fths of the overall voting population oppose intervention, but the majority, three-fths of the overall voting population, support it. The requirement of a supermajority of two-thirds of the voting population would mean that intervention would not occur. 23 International Committee of the Red Cross (in conjunction with Greenberg Research, Inc.), The People on War: ICRC Worldwide Consultation on the Rules of War (Geneva: ICRC, 2000), 54, from http://www.icrc.org/Web/eng/siteeng0.nsf/html/p0758 (accessed April 4, 2006). 24 Mary Kaldor, New and Old Wars: Organized Violence in a Global Era (Oxford: Polity Press, 1999), 121. 25 deLisle, Humanitarian Intervention, 552. 26 See, for instance, Michael Walzer, The Moral Standing of States: A Response to Four Critics, Philosophy and Public Affairs 9, no. 3 (1980): 20929. 27 Tesn, The Liberal Case, 107. 28 See Theodora-Ismene Gizeles and Kristin Kosek, Why Humanitarian Interventions Succeed or Fail: The Role of Local Participation, Cooperation and Conict 40, no. 4 (2005): 36383. 29 Kaldor, New and Old Wars, 126. 30 Jarat Chopra and Tanja Hohe, Participatory Intervention, Global Governance 10, no. 3 (2004): 291. 31 Gizeles and Kosek, Why Humanitarian Interventions Succeed or Fail, 36383. 32 Ibid., 364. 33 This concern about the burdens of intervention connects to the rst argument for increased effectiveness. If those subject to the intervention agree to its burdens, humanitarian intervention is more likely to be successful. But if these individuals oppose humanitarian intervention because of its burdens, the subsequent resistance is likely to make successful intervention difcult to achieve. 34 Tesn, Humanitarian Intervention, 126. 35 Tesn, The Liberal Case, 107. 36 These categories could be further divided. For instance, the category of accomplices could be divided into those who are willing, those who are naive, and those who have little choice but to be accomplices. But the typology (as presented above) captures the most morally relevant distinctions. 37 Strictly speaking, use of the term individual self-government may not always be appropriate for external representativeness because the intervener does not establish a government in the target state (unless it forms a transitional administration). Nevertheless, the underlying principle is essentially the same: individuals should have some degree of control over their ruling institutions. To show the mirroring of the argument for internal representativeness, I will continue to use the term individual self-government. 38 Albert Weale, Democracy (New York: St Martins Press, 1999). 39 Weale phrases this in terms of interests, but the argument can be applied to well-being. 40 It may be replied that, on an objective list view of well-being, we can dene the constituents of a good life and hence what is necessary for well-being. However, within the broad categories of the values that contribute to the good life (such as friendship), the details of the good life for each individual cannot be known a priori. The particular individual is the best judge of these details.

Representativeness and Humanitarian Intervention

41

587

Note that this is an argument for the intrinsic worth of individual self-government, not communal self-government (which is sometimes used to defend nonintervention). Some accounts of the value of communal self-government would include communities with little or no individual selfgovernment. See, for instance, Walzer, The Moral Standing of States, 20929. 42 One example of this was the response by a number of Somalis to the proposal to send Kenyan, Ethiopian, and Djiboutian peacekeepers to Somalia in 2005. Their opposition to intervention by their neighbors (especially Ethiopian) was so great that a brawl erupted in the Somali parliament and Somali warlords claimed they would target Ethiopian peacekeepers. BBC, Peacekeepers for Somalia Approved, March 18, 2005, from http://news.bbc.co.uk/1/hi/world/africa/4363021.stm (accessed June 27, 2006).